|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
24.99.22.29
In Reply to: RE: I like this one. posted by Tre' on November 17, 2014 at 07:30:07
If an R(k) has a parallel cap connected from cathode to ground, it is fully bypassed.
The amount of capacitance may be inadequate for full range performance, but that is a different issue.
Follow Ups:
"The amount of capacitance may be inadequate for full range performance......"Yes, that is what I meant.
When a cathode is bypassed to the point that the lowest frequency of interest is unaffected by the filter (amplitude and phase), I call that fully bypassed.
When the cathode is bypassed but the -3db point is higher than a decade below the lowest frequency of interest, I call that partially bypassed.
A fully bypassed cathode will prevent local current feedback from occurring down to the lowest frequency of interest.
A partially bypassed cathode will allow local current feedback to occur below a certain frequency depending on the value of the cap but not above that frequency.
This will give less gain in the lower frequencies and increase the plate resistance of the tube at those lower frequencies thus changing the damping factor of the amplifier if it's an output tube.
My use of the terms, fully bypassed and partially bypassed, can be found all over the internet. Rightly or wrongly.
BTW It is not the cathode resistor alone that is being bypassed but the cathode resistor in parallel with the cathode impedance, r.
The value of the impedance of the cathode of the tube is (the plate resistance + the plate load) / (the mu + 1)
That number is in parallel with the value of the cathode resistor to give us r.
r is what is what we are attempting to bypass.
This can make a huge difference. In the case of Dennis' amp it's not his 5k ohm cathode resistor that needs to be bypassed but that 5k ohm resistor in parallel with the 634 ohm cathode impedance. So 563 ohms is what needs to be bypassed.
56uf would get him 5Hz with 50Hz phase and amplitude correct.
141uf would get him 2Hz with 20Hz phase and amplitude correct.
Understand that I'm just talking about the effects of this filter, not the amplifier as a whole.
I believe Dennis uses 15uf and that will not fully bypass (by anyone's definition) 563 ohms.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 11/17/14 11/17/14 11/17/14
I am not attempting to be argumentative, but it does not matter what may or may not have become common usage on the internet; the words in use here have a very clear meaning.
Any cap (of any size) that is parallel to a given R(k) and connected to the cathode at one end and the ground at the other end comprises a fully bypassed R(k).
The effective bandwidth of such an setup is an entirely different matter; I would suggest calling it 'inadequately bypassed for full-range operation' (or something similar).
The cathode on Dennis' output tube is inadequately bypassed for full-range operation.
Is there a true meaning of "partially bypassed? Or is that just a made up term?
I think I heard that "vape" just became a real word.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
A partially bypassed R(k) is one in which some portion of the R(k) has a bypass cap and some portion of it does not.
To use a specific example (one that I am listening to as I type this): the cathode of the 5744WB submini has an R(k) totaling 627R - 600 ohms is bypassed by a 125 uf cap, and 27 ohms is left unbypassed.
While I agree that the unbypassed portion complicates design considerations...the combination of degenerative feedback, and what I theorize is some isolation from the cap sounds a lot better than simply bypassing the entire R(k).
Experiment, and let us know what you think.
Tre, Jim's idea is worth investigating.
I have been known to calculate the cathode bypass cap value by considering it separate from the tube cathode-- calculate it looking only at the cathode resistor-- just to have a look at THAT value independently.
I was defining "fully bypassed" as you were-- use the total tube cathode and resistor values to calculate the value you need for a certain low HZ point-- I don't use that much capacitance, but Jim has the DEFINITION right.
Mentally, it makes no difference how you view the definition as long as you understand what you're affecting, but if-- for instance-- I used your 56uf value-- as calculated in your earlier example's total cathode resistor value, a cap that large would badly affect the stage's H.F. response, not to mention that obtaining a high-quality cap of that value is also a problem. Of course, there's always Duelund....
Some cathodes are not bypassed at all, and the stage operates clean.
Adding cathode bypassing is an art. One should proceed cautiously, starting at the value calculated without the tube, slowly increasing the cap's value until you start killing the amp's best qualities-- definition, transparency, speed, etc.. While doing this I have my calculated cap value that I got from ignoring the tube cathode itself-- handy just for reference.
Too little bypassing will be evident as you know from your own experience.
Even a tiny bit too much capacitance will severely stifle the amp-- it will effectively muzzle it's best qualities, and it will sound like everybody else's amplifiers-- not bad, correct-- but not outstanding, not truly inspiring.
It's a process that requires the utmost care if you expect your amp to play with the Spectral DMA200S.
---Dennis---
"I have been known to calculate the cathode bypass cap value by considering it separate from the tube cathode-- calculate it looking only at the cathode resistor-- just to have a look at THAT value independently."
Well gee Dennis, that doesn't make any sense because, like it or not, the cathode impedance IS in parallel with the cathode resistor.
Bypassing just the cathode resistor value will leave you with too small a cap.
"Too little bypassing will be evident as you know from your own experience."
Too little bypassing will increase the plate resistance of the output tube and leave you with less damping factor.
To my way of thinking, 56uf in your output stage is a minimum value. A value that will only "bypass" the cathode to a -3db point of 5Hz meaning there will be some current feedback all the way up to 50Hz, phase shift also.
I'm sure one can get away with cathode bypass cap that are too small in voltage amplifiers, to one extent or another, but output stages are different.
The relationship between the output tube's plate resistance and output transformer's primary impedance is critical and when the cathode bypass cap is too small all hell breaks loose.
You should read a little Lynn Olson on this subject.
(In effect, the performance of the tube is greatly degraded — exactly the same effect as a large drop in emission.)
Dennis, I can't tell you what type of sound you should like but I can tell you that when the cathode bypass cap on the output tube is too small the amplifier will not be right.
Your speakers might "like" the lack of damping and the lower amplitude in the LF caused by degenerative feedback because the bypass cap is too small, but that's a different matter.
BTW If you are worried about the quality of the sound, at the high frequencies, of larger caps you can use a smaller cap value and still have the cathode bypassed all the way down to the lowest frequency of interest by using Jack Elliano's UltraPath technique.
This also helps to "take the power supply" partially "out of the signal path".
This is what I do and I can tell you it's the way to go. If you haven't heard my amps, if you haven't taken the time to come to my house and hear my amps then you are just guessing. Everyone who has not come to my house has never heard it and you don't get-it.
You have never heard my amps, in many venues, as I have, have you??
You are talking unsubstantiated, only guessing, with zero practical experience.
Have YOU ever heard a pair of my amps???
So the facts remain : (1) you never heard a Tre Perry amp, neither of the people, and how well it plays and (2) I have taken the basic amp to a new and even higher performing level with my latest implementations, same circuit, better implementation, the best possible !!
OK, that's enough of that. :-)
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Doing a DF on DF!!! - I gotta take my hat off to ya Tre' ....
Edits: 11/19/14 11/20/14
I thought it was funny but DF didn't get it.
Oh well.
I guess I'm not much of a comedian.
But Seriously, what if each of us just flatly stated that our way of doing things was the only way.......even though, from a technical point of view, our way was the wrong way?
That just sounds like anarchy to me.
In any field, especially a technical one like audio electronics, humans build on existing knowledge. But instead of that, Dennis just seems to disregard existing knowledge, even holding it in disdain.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
I was also unaware that you were attempting to be humorous - but, I'm a nerd and often miss such things...this is not a personal attack; however, you appear to have a vendetta against Dennis Fraker and/or Serious Stereo. Why is that? Do you own one of his amps?Have you heard his amps? Was it so horrible that you felt you must spend hours protecting the public?
Has he screwed you (or family members) over?Are you expecting Dennis to apologize for his design decisions, and beg for your forgiveness/engineering assistance? If he did so, would his amplifiers then be acceptable?
Have you attacked Audio Research, Jadis, VTL, or Conrad-Johnson for their many design compromises?
What are you looking for...or expecting?
Edits: 11/20/14
But what can people measure in wires (if they are carrying DC), with any kind of equipment, other than voltage, current and impedance, and if one can measure anything else, do many of them conduct these measurements on each strip of wire they lay down between two components?I know in some cases one can hear differences by changing wires? But can one see differences you hear?
Observe, before you think. Think before you open your yap. Act on the basis of experience.
Edits: 11/20/14
I'm delighted that your amp sounds good-- no offense meant.Damping factor-- I think I got that "perfect". It plays a lot of speakers that it shouldn't, and stays clean and solid right down into some unbelievable lows.
The interesting is that H.F. response is very extended-- enough to place performers right there and fool you. Cymbal clashes, and many other H.F. events sound real, present, and full, and extended.
You're not going to be able to "explain" that, although your attempts are certainly entertaining at the least.
"All Hell breaks loose". HUH? I probably have the most stable amp ever built! NOTHING fazes it! NOT ONE has ever lost a power tube or a rectifier.
Drivers? One could blow a grid out of one by putting a huge voltage spike straight into the input. I tried this a few times on tests. Nothing happened at all, just silence. NOTHING happened. I had to replace the driver to get music!
"MY" speakers? People play these amps on all kinds of things....
"Zero" practical experience? Oh my gosh! Amazing!
Tre, I'd love to hear your amps! In fact, if I make it to RMAF once again, you can bring them in. After hours, we'll hook them up. If they're better than mine, we'll just leave them in (with your permission) and just let the visitors listen to those!
On another subject, scientists at NASA think we've entered a Sun-Spot minimum period, a period of substantial Earth cooling. (Sort of like the old "Maunder Minimum", etc.).Russian scientists are a bit bolder, some flatly state that we've already entered a new "Ice-Age"-- a 30-50 year one, not a major one--, but one that's going to cut food production way back, and will make your Winter energy bill soar.
If this is true, yours truly-- and perhaps some of you as well-- will be engineering something other than audio amplifiers.
---Dennis---
!
Edits: 11/19/14
"You're not going to be able to "explain" that..."That's because there is no explaining it. It isn't real, just your imagination.
""Zero" practical experience? Oh my gosh! Amazing!"
That was my attempt at sarcasm. These are the kind of things others say when they no longer have a technical argument.
If you want to hear my amplifiers you will have to come to my house.
""MY" speakers? People play these amps on all kinds of things...."
How many people? Dennis, how many pairs have you sold? 5 maybe?
Be honest.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 11/19/14
"...an R(k) totaling 627R - 600 ohms is bypassed by a 125 uf cap, and 27 ohms is left unbypassed. "
Oh, I see.
Thanks.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
A lot of the old school split load phase inverters seem to use this method, remember? It even shows up in some guitar amps.
I think there's 2 "partially bypassed" versions.
one is less than full bandwidth, but across the entire Rk
the other is across part of the Rk accommodating whatever bandwidth,
with whatever amount of FB.
How's about 'partial bypass' and 'sub critical bypass' ?
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: